December 6, 2007
Medium-Temperature Pressure Leaching of Copper Concentrates – Part IV: Application at Morenci, Arizona
By Marsden, J O Wilmot, J C; Smith, R J
Abstract Parts I and II in this series of papers presented the chemistry of medium-temperature pressure leaching of copper sulfide concentrates and reviewed the metallurgical development of a process to effectively recover copper using medium-temperature pressure leaching and direct electrowinning (DEW). Part III reviewed the large-scale demonstration of the MT-DEW-SX process at Bagdad, Arizona. Key features of the process were the use of superfine grinding of the concentrate (to a P^sub 98^ of 12 to 15 [mu]m and a P^sub 80^ of 6 to 7 [mu]m), pressure leaching at 160[degrees]C with 1,380 kPa (200 psig) oxygen overpressure, DEW to produce LME Grade A cathode, recycling some of the lean electrolyte to the pressure leaching step, and treatment of a lean electrolyte bleed by solution extraction (SX). Recycling of lean electrolyte provided all of the acid required to effectively dissolve copper and convert approximately 65% to 70% of the sulfide sulfur to elemental sulfur, resulting in an acid-autogenous process for pressure leaching. This paper (Part IV) discusses the development and design of a full- scale commercial application for the Morenci operation in Arizona.
Key words: Copper, Pressure leaching
Introduction and background
Part I in this series of papers presented the chemistry of medium- temperature pressure leaching of copper sulfide concentrates and reviewed the initial metallurgical development work. Part II provided details for the testing and development of an integrated copper extraction process that included superfine grinding of concentrates to 98% minus 12 to 15 [mu]m (80% minus 6 to 7 [mu]m), medium temperature (MT) pressure leaching at 160[degrees]C with 1,380 kPa (200 psig) oxygen overpressure, direct electrowinning (DEW) to produce LME Grade A cathode, recycling of the lean electrolyte to the pressure leaching step and treatment of a lean electrolyte bleed stream by solution extraction (SX).
In the third quarter of 2003, continuous pilot plant campaigns were conducted at Hazen Research Inc. using the MT-DEW-SX flowsheet to treat concentrates from Bagdad (Arizona) and Cerro Verde (Arequipa, Peru). The copper mineralogy in the concentrates from both locations was primarily chalcopyrite. This work successfully demonstrated the process and indicated copper extractions of 97.5% from the Bagdad material and 96% to 97% for the Cerro Verde material (Marsden et al., 2007b). In late 2003, a decision was made to convert the Bagdad hightemperature leaching plant to operate in the MT-DEW-SX configuration to allow the process to be demonstrated at a small commercial scale, ~12,700-t/a (14,000-stpy).
The conversion of the Bagdad facility was completed in early 2005, and the plant was operated in MT-DEW-SX mode for a period of seven months. Direct electrowinning (DEW) successfully produced LMEGradeAcathode. The process was demonstrated to be acid- autogenous (i.e., no concentrated sulfuric acid was required as make up to the pressure leaching feed), and it generated some acid that could be used beneficially in heap and/or stockpile leaching. The medium-temperature pressure-leaching step converted 65% to 70% of the sulfide sulfur in feed to elemental sulfur, generating approximately one-third of the amount of sulfuric acid compared with the hightemperature process. This work was reported in Part III of this series of papers (Marsden et al., 2007c). After completing the demonstration testing at Bagdad in late 2005, the MT-DEW-SX process was considered to be ready for large-scale commercial application.
It should be noted that throughout this paper the terms "solution extraction" and "SX" are used to refer to the process that is commonly referred to as "solvent extraction" in the industry.
Morenci District developments
In early 2004, the potential application of concentrate-leaching technology to process chalcopyrite-dominant material from the Western Copper deposit in the Morenci district was recognized. At that time, planning was under way to restart the Morenci concentrator, which had been idled since 2001. The Morenci concentrator would treat the chalcopyrite-dominant material to be mined out of the Western Copper deposit starting in about 2007 and would produce a flotation concentrate that would be shipped to the Miami smelter in Miami, Arizona, for subsequent treatment by conventional smelting and refining.
A preliminary feasibility study for the application of concentrate leaching at Morenci was prepared by Aker-Kvaerner, including a trade-off study between high- and medium-temperature pressure leaching. At the time of the study, the existing Morenci operations consisted of crushing and heap leaching of the high- grade portion of chalcocite-dominant ore and run-of-mine (ROM) leaching of the lower- grade chalcocitedominant material. Copper was subsequently recovered by four SX plants and three EW facilities at a total production rate of approximately 400,000 t/a (360,000 stpy) of cathode copper. The leaching operations required a significant amount of sulfuric acid, most of which was supplied from the Miami smelter. The sulfuric acid was added to the agglomeration drums in the crushing and heap-leaching circuit, was added into the EW electrolyte as make up for the acid lost in the electrolyte bleed solution and was added into the heap and ROM stockpile leach raffinate solutions. Concentrate leaching provided an opportunity to supply a portion of the acid requirements directly at site. In addition, it was projected that there would be approximately 75,000 t/a (68,000 stpy) of EW copper-plating capacity available at Morenci from mid-2007 onwards.
In late 2004, following the completion of the preliminary feasibility study, the decision was made to proceed with continuous pilot plant testing at Hazen Research to develop a process flowsheet for this application. Based on the results of a trade-off study and acid-dosing requirements forecasted at site, the MT-DEW-SX flowsheet provided a better fit with the Morenci application than high- temperature pressure leaching followed by SX-EW. Several "surrogate" concentrate samples that were expected to have similar composition (both chemistry and mineralogy) to the Western Copper concentrate were prepared and used for the continuous pilot plant work. This was necessary because sufficient representative bulk concentrate was not available from Western Copper and could not easily be obtained in time to support the proposed project development schedule. However, the project team was confident that the surrogate concentrate materials would be adequate to support the commercial plant design based on the following factors:
* detailed chemical and mineralogical information was available for samples throughout the Western Copper deposit;
* three major ore types were identified within the Western Copper deposit, and the expected flotation concentrate composition for each ore type could be predicted based on previous metallurgical work in the district;
* earlier continuous pilot plant work at Hazen Research had tracked impurities in the solution phases and final cathode product, and potential "bad actors" had been identified (no fatal-flaw elements were observed in the Western Copper materials); and
* the process was expected to be able to accommodate reasonable variations in feed mineralogy and would be designed accordingly.
Based on this assessment, a laboratory program was initiated at Hazen Research to determine the pressure-leaching operating conditions necessary to obtain high copper extractions with different sulfide mineral compositions, as well as provide essential design criteria for the preferred process. This work is reviewed in detail in the following sections.
Continuous pilot work on surrogate Morenci concentrate samples
A comprehensive pilot plant program was developed for the Morenci application with the following objectives:
* determine the process-design criteria, including preferred operating conditions, for the MT-DEW-SX process at Morenci with three different copper sulfide mineral blends (although partially composed of flotation concentrates from operations not related to Morenci, these blends were expected to represent the mineral assemblage that would be fed to the commercial Morenci pressure leaching circuit);
* determine the kinetics of sulfide oxidation and copper extraction in the pressure leaching circuit and measure iron dissolution in pressure leaching discharge slurries over time;
* perform mineralogical evaluations on the residues to further understand the pressure-oxidation reactions;
* determine scale buildup and S[degrees] accumulation within the pressure-leaching vessel and quantify the sulfur deposition in the gas vent system;
* monitor the impurities from the oxidation operation and evaluate their effects on EW;
* determine the neutralization kinetics for a simulated countercurrent decantation (CCD) underflow; and
* collect solid, liquid and gas samples for environmental analyses.
Surrogate Morenci concentrate composition of pilot plant feed
Based on an internal evaluation of the Western Copper deposit at Morenci, three concentrate compositions (Ore Types 4,7 and 10) were selected by Phelps Dodge for pilot testing. (Phelps Dodge is a wholly owned subsidiary of Freeport-McMoRan Copper & Gold Inc. following a merger on March 19, 2007.) These ore types contained various ratios of copper sulfide minerals and pyrite (Table 1) and were believed to represent the expected range of minerals that would be fed to the commercial pressure-leaching circuit over the lifetime of the project. Because the concentrator was not yet running, actual Morenci concentrates with these compositions were not available. Therefore, pilot feed blends were prepared using copper concentrates from the Morenci (Arizona), Chino (New Mexico) and Bagdad (Arizona) operations, plus a separate pyrite concentrate. The Morenci concentrate was generated from a small pilot plant at Morenci to provide a covellite-rich fraction for blending. This concentrate was further upgraded by flotation at Hazen Research to remove excess gangue. The Chino concentrate was also upgraded at Hazen Research by flotation to remove pyrite, providing chalcocite-rich concentrate for blending. The bulk of the chalcopyrite used for blending came from as-received Bagdad concentrate. These four sources provided the mix of sulfide minerals needed to prepare the three ore types. Table 1 provides the approximate mineral composition for each of the ore types used in the pilot plant, based on optical microscopy. Because of the excessive amount of flotation reagents used to prepare some of the mineral concentrates for the surrogate blends, there was some concern that this might adversely affect copper extraction in the pilot pressure-oxidation vessel. This factor was considered carefully in the analysis of pilot plant results. Table 1 - Approximate mineral and chemical compositions of surrogate pilot plant Morenci concentrates.
Pilot plant description
During January and February 2005, a total of 19 pilot plant runs were conducted on three samples of concentrate that were generated to simulate material to be produced from the three major ore types in the Western Copper deposit at Morenci, Arizona. Three additional runs were conducted using Candelaria chalcopyrite concentrate for comparative purposes. The duration of each run typically varied from 9 to 12 hours, depending on the operating conditions and stability. Prior to these tests, four initial pilot plant runs were conducted to build up the copper, iron and acid concentrations of the pressure oxidation discharge solution close to equilibrium levels for the closed loop system. In general, the flowsheet, equipment and base operating conditions were similar to those used in previous MT-DEW- SX campaigns (Marsden et al., 2007b), except as noted below. The general flowsheet for the pressure-leaching pilot plant is shown in Fig. 1.
For each pilot plant run, the sample of concentrate to be tested in each case was subjected to superfine grinding to a 98 % passing 12 urn target size followed by pressure leaching at 160[degrees]C for 90 minutes with continuous oxygen sparging (1,380 kPa oxygen overpressure). Calcium lignosulfonate (CLS) was added to the feed at a dosage rate of 10-kg/t concentrate. Other operating parameters were established from batch test results on various blends of sulfide minerals. The purpose of the pressure-leaching tests was to optimize copper extraction and maximize elemental sulfur make, while achieving acceptable iron concentration in the discharge solution to support DEW downstream (Marsden et al., 2007b).
The pressure leach discharge - containing copper, iron and acid in solution and a mixture of hematite, elemental sulfur and iron sulfo-salts in the solid residue - was subjected to solid-liquid separation. The copper-rich solution was further clarified (filtered) and fed directly to EW for copper recovery. The cooling liquor for the pressure leach vessel was provided by a combination of clarified pressure-leach discharge (i.e., strong PLS), CCD overflow solution (i.e., weak PLS) and lean electrolyte from the Morenci operation. This was done to ensure that the solution recycled to the pressure-leach vessel represented the commercial electrolyte as closely as possible, while ensuring that the required acid addition was provided to the pressure-oxidation process.
Superfine grinding. The concentrate samples prepared by flotation were stored under water until required. As needed, the samples were ground in a Metso Minerals 7.5-kW Stirred Media Detritor (Metso SMD) using preconditioned 2.0 x 3.4 mm (6x9 mesh) Colorado River Sand obtained from Oglebay Norton's Colorado Springs plant as the grinding media. The Metso SMD mill was used for this purpose (as opposed to the Netzsch Model LME4 mill used in the previous medium- temperature pilot campaigns at Hazen Research) because Metso SMD mills were in the process of being installed at Bagdad for the medium-temperature demonstration at the time of conducting the Morenci pilot program. (Note that the Morenci operation is designed to utilize a single IsaMill M10000 unit in the commercial plant.) Grinding was conducted at approximately 55% solids. A Malvern Mastersizer particle size analyzer was used to provide particle size information for the product.
Pressure leaching. The pressure-leach vessel and the associated equipment used for this purpose were identical to that used in previous medium-temperature continuous pilot plant work, as reported previously (Marsden et al., 2007b). One important exception to the pressure-leaching operation was that, during the four initial pilot plant runs, the pressure leach vessel agitators were operated at a speed of 600 rpm, but the speed was increased to 800 rpm for the remainder of the pilot runs to ensure effective oxygen mass transfer.
For each of the three concentrate types, three runs were conducted to investigate the effect of acid addition (high, medium and low) on copper extraction and conversion of sulfide sulfur to elemental sulfur (the sulfur "make"). These were followed by a few runs to test other parameters, including increased residence time, increased calcium lignosulfonate addition, finer grind size, higher and lower temperatures, and the addition of tannin. For each concentrate type, a run was performed at the "optimal" conditions to provide data for process design.
Commercial-scale pressure-leach circuit systems to treat sulfide concentrates operate in an energy-autogenous mode, and the operating conditions are adjusted to maintain the desired operating temperature without the requirement of external heating or cooling where possible. Because of its small size, the pilot pressure leach vessel at Hazen Research Inc. had significant heat losses compared with those expected for a commercial unit. Consequently, rather than operating the pilot unit in an energy-autogenous mode, an energy balance was used to set the pilot plant flow rates to simulate commercial operation. This required that the pilot-scale vessel be externally heated to compensate for the higher heat losses. A heat- balance program was developed for this purpose. The program considered a number of input variables, including vessel working volume, concentrate analysis, target retention time, operating temperature, oxygen overpressure, acid addition rate, acid concentration in solution, slurry feed solids content and cooling solution concentrations. The heat-balance generated target conditions for concentrate slurry feed rate, acid feed rate, approximate oxygen flow rates and cooling solution flow rates into each pressure leach vessel compartment.
Figure 1 - Circuit flowsheet for the pilot pressure leaching Plant at Hazen Research Inc.
To compensate for the lack of a flash pressure let-down system and the evaporation that occurs during flash let-down, the pilot unit was fed at a higher feed-to-cooling solution ratio than predicted by the heat and mass balances. This was done to generate discharge solution with species concentrations similar to those expected from a commercial pressure-leach circuit.
Specific samples were collected during the operation to provide important additional design data. Pressure-leach discharge slurry samples were collected to perform solid-liquid separation tests for sizing the primary thickener, the CCD circuit and the thickener underflow filter. Gas samples were collected from the off-gas vapor for each of the concentrate sample types for environmental analyses. At the end of the process design runs at "optimal" conditions, the off-gas vent lines were flushed with xylene to quantify the amount of elemental sulfur that was collected during operation. Also, during the design runs, residue samples were collected for mineralogical evaluation to assist in better understanding the pressure-leaching chemistry within the vessel for each concentrate type.
For a few of the runs, pressure-leach feed and product samples were collected and analyzed for a suite of elements to assess potential impurity problems in the DEW process downstream.
Direct electrowinning (DEW). Comprehensive pilot testing and development for the integrated operation of medium-temperature pressure leaching, DEW and SX treatment of lean electrolyte bleed (plus low-grade solution) was completed in 2003, and this was not repeated in the Morenci pilot campaign. In addition, the large- scale demonstration of the MT-DEW-SX flowsheet was in the final stages of construction at Bagdad (Arizona) at the time of the Morenci testing. Attention was focused on impurity identification, quantification and deportment, with emphasis on elements that could have an effect on ultimate cathode quality.
The flowsheet that was developed for the Morenci application out of this work included the use of second- and third-stage electrowinning (referred to as "EW2" and "EW3," respectively) to remove copper from the lean electrolyte bleed as required by Morenci electrolyte balances. Design of the system was to produce a solution containing sulfuric acid (245 g/L) but little copper (5 g/L), which could be directed to the heap and/or stockpile leach systems for subsequent treatment by SX. This would allow the acid value of the pressure-leach solution to be utilized in heap and stockpile leaching operations. Figure 2 - Conceptual flowsheet of EW2 and EW3 for Morenci pilot plant runs.
The lean electrolyte bleed solution at Morenci was projected to be approximately 240 m3/hr containing 34 g/L copper, 200 g/L free sulfuric acid and about 2 g/L iron. The concept for the use of second- and third-stage EW is shown in Fig. 2. Independent of the pressure-leaching pilot plant work described above, a pilot program was developed to test the electrolyte bleed treatment scheme. The preliminary process design criteria for the EW2-EW3 steps are included in Fig. 2 and are not reported further here.
Results of pilot plant campaign
The results of the pressure-leaching pilot plant design basis runs for each concentrate type are summarized in Table 2. The pilot plant campaign indicated copper extractions of 96.4% for two of the surrogate concentrate types (representing Ore Types 4 and 7 from Morenci) and 94.0% for the third surrogate concentrate material (representing Ore Type 10). For Ore Types 4 and 7, the optimal acid concentration was found to be 250 to 300 kg/t. For Ore Type 10, copper extraction was essentially independent of acid concentration. The conversion of sulfide sulfur to elemental sulfur varied from approximately 40% for Ore Types 4 and 7 to more than 50% for Ore Type 10.
Table2-Pilot pressure oxidation resultsforthe design runs with Morenci pilot plant concentrates. (The stated results were averaged from the design pilot plant runs for samples of each ore type using the conditions selected forthe process design criteria.)
Superfine grinding results. Based on the limitations of the pilot grinding equipment used, a particle size of 98% passing 13 to 17 um was targeted for these runs. However, the various mineral concentrates used to make up the blended samples were observed to have quite different hardness, resulting in wide variations in the ground product size distribution. For most of the runs, the concentrate slurry was fed through the 7.5-kW Metso SMD mill in two passes at a flow rate of 1.2 to 1.8 kg/min. For two of the runs, a single pass was used, because the material was found to be significantly softer than the average. The nominal operating parameters for the mill were 6.0 kW power draw and a product exit temperature of 74[degrees] to 78[degrees]C (~66 kWh/t power application). To prevent oxication, the superfinely ground concentrate was held under water in drums until required for the pressure-leaching campaigns.
As shown in Fig. 3, the power required to produce the appropriate final grind size was strongly dependant on the individual concentrate hardness.
Pressure-leaching results. Copper extraction as a function of acid addition for each ore type showed some scatter, but generally demonstrated a slight decrease in copper extraction with higher acid addition for Ore Type 4, a larger decrease for Ore Type 7 and no change for Ore Type 10. This indicated that acid additions in the low-to-intermediate range were the most favorable for copper extractions for Ore Types 4 and 7. Ore Type 10 was not affected by changes in acid addition between 300 and 500 kg/t. All acid additions in this data set included the contribution from iron in the cooling liquor, as discussed in Part I of this paper series (i.e., hydrolysis of Fe^sup 3+^ to hematite, releasing sulfuric acid in the pressure-leach vessel).
Figure 3 - Grinding work input (kWh/t) vs. particle size (P^sub 80^ in [mu]m).
Figure 4 - Acid addition vs. copper extraction for the surrogate Morenci concentrates.
Figure 4 illustrates the effect of acid addition on copper extraction from the pilot runs for the three Morenci surrogate concentrates. The amount of acid required for the highest copper extraction for the three ore types was shown to be less than that required for the Bagdad and Cerro Verde chalcopyrite concentrates studied previously (400 to 450 kg/t vs. 500 to 600 kg/t). This was thought to be due to the higher pyrite content of the surrogate Morenci ore types (19% to 23% pyrite) compared to the Bagdad and Cerro Verde concentrates (5% to 8% pyrite). Based on the results of the laboratory batch tests, 6% pyrite in the feed was equivalent to 100 kg/t acid addition.
The significant effect of pyrite concentration in the feed on the amount of acid required for the pressure-leach step reiterates the importance of removing the bulk of the pyrite from the concentrate during flotation. However, overdosing of flotation reagents may also be detrimental to copper extraction as continuous tests have shown previously (Marsden et al., 2007b). The presence of chalcocite and covellite in Ore Type 7 did not lower the copper extraction compared to Ore Type 4; however, for Ore Type 10, the additional chalcocite decreased the copper extraction by 2.4% (from 96.4% to 94.0%). The higher chalcocite content (36%) in Ore Type 10 appeared to have negatively influenced the extraction of copper, but the presence of a moderate concentration of chalcocite (22%) in Ore Type 7 did not affect copper extraction. Mineralogical evaluation indicated that covellite showed a tendency to become encapsulated by S^sub o^ and resistant to further oxidation in the pressure leach vessel. Consequently, it was determined that native covellite in the feed, or covellite formed from chalcocite in the feed, has the potential to become encapsulated, resulting in lower copper extraction. Table 3 shows the results of X-ray diffraction analyses performed on compartment samples for Ore Type 7 material. Clearly, the covellite did not react as quickly as the chalcopyrite, pyrite or chalcocite.
In addition to affecting the copper extraction in the vessel, the total acid addition (including the contribution from Fe^sup 3+^ hydrolysis) also impacts the discharge acid and iron concentrations in solution. As one would expect, increasing the total addition of acid to the leaching vessel increases the discharge acid level concentration. However, this increased acid addition results in a corresponding increase in the iron concentration in solution. Figure 5 illustrates the relationship between the discharge solution iron concentration and the discharge solution acid concentration.
When handling the pressure-leach discharge slurry samples from the pilot runs, it was observed that the iron concentration in solution was variable over time, typically increasing as the solutions were allowed to stand in contact with the hematite-sulfur residue. To quantify this behavior, a series of tests were run in which the discharge slurry from a particular pressure-leaching run was allowed to remain in contact with the residue at temperatures that one would expect to encounter in a typical operation. The data are shown in Fig. 6, where iron concentration in solution is plotted as a function of time in contact with the residue. The results show a variation in solution iron response, depending on pressure-leach test runs, but all show a definite increase in iron over time. This highlights the need for a thorough and rapid separation of the hematitesulfur residue from the strong pregnant leach solution as iron levels above approximately 2.5 g/L can have a negative impact on the current efficiency in the tank house.
Table 3 - X-Ray diffraction analyses of Run 16 profile samples for ore Type 7.
Figure 5 - Relationship between discharge solution acid concentration and iron concentration.
Figure 6 - Solution iron concentration as a function of residue contact time.
Solid-liquid separation. Three pressure-leaching discharge slurry samples representing the three surrogate Morenci ore types were collected during the pilot plant operation for solid-liquid separation tests by Pocock (Salt Lake City, Utah). These tests were run soon after collecting the samples to minimize aging. The slurry samples were taken from three runs representing Ore Types 4, 7 and 10.
A range of solid-liquid separation tests were completed, including:
* flocculant screening,
* static thickening for conventional thickener and CCD circuit design,
* dynamic thickening for high-rate thickener design,
* viscosity of thickened pulp for rake and underflow piping design,
* pinned-bed clarification of overflow solutions,
* vacuum filtration for belt vacuum filter sizing and
* pressure filtrationforpressure-filtrationequipmentdesign and sizing.
A medium-to-high molecular weight non-ionic polyacrylamide flocculant provided the best overall performance. For the decant thickener, Ore Types 4 and 10 produced good overflow clarities and thickening using 60 to 70 g/t flocculant, but Ore Type 7 required 100 to 120 g/t. All CCD tests produced good thickening results within a feed solids concentration range of 5% to 10%. For the decant thickener, total suspended solids (TSS) in the overflow ranged from 400 to 600 mg/L. Generally, the CCD produced overflows with lower suspended solids.
Table 4 - Gravity sedimentation-static thickening results.
The results of the pinned-bed clarification tests showed that solids were removed down to the detection limit. At the completion of the test, the solids reached acceptable densities of between 12% and 15% (by weight). The results of the settling tests are presented in Table 4.
The rheological behavior of the thickened pulps across a specific shear rate range was examined. All decant thickener simulation underflow samples showed Bingham plastic characteristics, and after yield values were reached, the pulps behaved as near-Newtonian fluids. Ore Type 4 underflow slurry exhibited a yield value near 30 Pa at a solids concentration of about 45%. Yield values for Ore Types 7 and 10 underflow slurry approached 30 Pa at about 40% solids. A series of tests to size the vacuum filter was conducted at ambient pH, at 60[degrees]C and at an applied vacuum of 67.7 kPa. A filter aid was required to generate a reasonably clear filtrate and easily discharged cake within reasonable drying times. All cakes were about 15 mm thick. Pressure filtration tests examined the effect of cake thickness and air-dry duration on production rate and filter cake moisture for the thickener underflow samples. Tests were conducted at 60[degrees]C with an overpressure of 552 kPa (80 psi). Production rates for an automatic recessed plate filter and a Larox- type pressure filter were estimated for equipment sizing. All filter cakes appeared to be stackable, and the filtrate was clear.
EW2 and EW3 development. Copper was recovered from the Morenci lean electrolyte bleed by EW to produce copper cathodes and lean electrolyte low in copper and high in acid. To achieve a compromise between the quality of cathode copper, the number of cells and the lowest possible copper concentration in the final lean electrolyte, a staged EW2 and EW3 tank house arrangement was considered, as shown in Fig. 2. Pilot testing of EW2 was relatively straightforward and was conducted in a similar manner to other EW pilot testing at Hazen Research. The method was reported previously (Marsden et al., 2007b) and is not reported further here.
To further develop the EW3 step, copper was plated continuously at EW3 conditions on a stainless steel blank in the pilot cell, producing 600 to 750 g Cu/day, depending on the current setting. At this rate, the copper cathodes grew approximately 1.2 mm in thickness per day (provided a smooth plate was deposited). Three cathodes were prepared in March and April 2005, and deposition times varied from 3 to 10 days. Cathodes were checked for appearance while interrupting the power and lifting the cathode out of the cell once during the plating. The anodes used in the pilot plant were Pb-Ca- Sn anodes. The copper plates were stripped manually after slightly flexing the cathode.
To generate the 5-g/L Cu electrolyte needed for EW3, the pilot EW circuit was filled with Morenci lean electrolyte. Then, the copper concentration was reduced to 5 g/L by EW. In previous MT-DEW pilot work at Hazen Research in 2003, copper was electrowon at EW2 conditions of 150 A/m^sup 2^; the lean electrolyte contained 19 g/L Cu, 170 g/L H^sub 2^S0^sub 4^ and between 2.5 and 3.5 g/L Fe; and the temperature was 50[degrees]C. Most of the cathodes met chemical specifications for LME Grade A. To determine the cathode quality and the current efficiency for EW3, a pilot plant was operated at the conditions specified, i.e., ~250 g/L H^sub 2^S0^sub 4^ and 5 g/L Cu.
In summary, cathodes made at the EW3 conditions had a granular appearance. Although the cathodes stripped easily manually, they did not appear to have sufficient strength to be handled in a commercial stripping machine. The current efficiency was ~70%. Of the six samples analyzed, five met the chemical specifications for LME Grade A, and one sample was out of compliance with respect to sulfur (45 ppm). Further EW2 and EW3 piloting at longer plating cycles resulted in cathodes that were more amenable to mechanical stripping and design criteria for these operations have been developed for this flowsheet option.
Process design for the Morenci application
Aker-Kvaerner Inc. completed a feasibility study in May 2005 for the application of the MT-DEW-SX process at Morenci to treat flotation concentrate from Western Copper. In June 2005 the decision was made to proceed with detailed engineering and preliminary construction activities at Morenci in anticipation of successful completion of the large-scale demonstration of the MT-DEW-SX process at Bagdad. In November 2005, a detailed review and fatal flaw analysis was completed for the Bagdad technology demonstration, and the decision was made to proceed with full construction of the facilities in Morenci.
The Morenci MT-DEW-SX process flowsheet is shown in Fig. 7. The major differences from the MT-DEW-SX flowsheet that was demonstrated at Bagdad are summarized as follows:
* A residue filtration step was included immediately following decant thickening to minimize the amount of time that the strong PLS would remain in contact with the hematite-sulfur solid residue from pressure leaching. This was done to minimize the iron concentration in the strong PLS and to ensure that the iron concentration was maintained at or below 2.5 g/L in the feed to direct electrowinning. Filtration also has the effect of increasing the amount of copper available for direct electrowinning.
Figure 7 - Morenci concentrate leach process flowsheet.
* A step was included to remove silica from the strong PLS to control the amount of silicon in the electrolyte. This was considered to be necessary to minimize the potential to form crud in the SX stripping stages and to limit the concentration of silicon in the final cathode. The silica removal process is proprietary Freeport-McMoRan technology.
* The direct electrowinning (treating strong PLS produced by pressure leaching) and conventional electrowinning (treating rich electrolyte produced by SX) steps were combined by combining the strong PLS and the SX rich electrolyte. This was considered to be a desirable change because the electrolyte composition would more closely resemble the conventional EW electrolyte compared with the DEW electrolyte treated at Bagdad in the demonstration plant.
In addition to these changes, the flowsheet that was considered in the Hazen pilot plant work on surrogate Morenci concentrate materials (and in the feasibility study) included two-stage electrowinning ("EW2" and "EW3") to treat the lean electrolyte bleed. During detailed engineering of the Morenci facility, it was determined that the EW capacity that was to be used for the EW2 and EW3 steps could be better utilized for increasing copper production identified in the mine plan. Consequently, the decision was made to take the lean electrolyte bleed directly to the heap/stockpile PLS for subsequent treatment by SX. This represented an additional cost to handle the copper in the bleed, but this was the best alternative.
Morenci process description. Production from the Morenci run-of- mine stockpile leach and crushed ore heap leach ("mine-for-leach") operations was expected to decline by mid-2007, so that the copper processed through the MT-DEW-SX circuit could be recovered without the need for expansion of the existing EW facilities. Additionally, the acid produced by pressure leaching would be used to largely offset concentrated sulfuric acid that is currently, and would be in the future, imported from the Miami smelter. General process design criteria for the Morenci Concentrate Leach Plant are shown in Table 5.
The design feed rate to the pressure-leaching circuit was determined to be 29.1 t/h (32.1 stph), based on 85% availability and 97.5% copper recovery, yielding copper production of 66,000 t/a (73,000 stpy) of copper and 384 t/day (423 stpd) of acid equivalent in the strong PLS solution. This production rate closely matched the average concentrate production expected from the Morenci Concentrator when processing ore from the Western Copper deposit, resulting in minimal transport of concentrate to the Miami smelter.
The concentrate is to be reclaimed by front-end loader from the existing bedding plant, which will be used for concentrate storage. The filtered concentrate will be repulped with fresh water and reground to a P^sub 80^ of 7 [mu]m (and P^sub 98^ of 15 [mu]m) in a superfine grinding mill (IsaMill M10000 from XstrataTechnology, Brisbane, Australia) using ceramic grinding media. The ground concentrate slurry will be fed to two pressure-leach vessels using high-pressure diaphragm pumps.
The pressure leach vessels are 4.3-m- (14.2-ft-) diameter by 25.2- m- (82.5-ft-) long horizontal steel pressure reactors, supplied by Eaton Metal Products Co., Salt Lake City, Utah. The vessels are lined with two courses of 76-mm- (3-in.-) thick acid-resistant brick on top of a Pyroflex(R) corrosion-protection membrane (full lining system provided by Koch Knight LLC, East Canton, Ohio). The vessels and lining system are designed to operate at a pressure of 1,918 kPa (278 psi), with a maximum rating of 2,242 kPa (325 psi). Each vessel is divided into six compartments equipped with six Lightnin A-340 agitators (Lightnin, Rochester, New York) fitted with mechanical seals. Gaseous oxygen will be injected below each agitator through bottom entry sparge tubes. The reaction is thermally autogenous and requires no external heat. The slurry will be discharged from each vessel, under pressure, through a choke system into a single-stage flash vessel (one for each pressure leach vessel). The flashed steam will be cleaned in a two-stage venturi scrubber system prior to discharge to the atmosphere. Most of the steam will be condensed using stockpile leach PLS or raffinate solution, which reduces air emissions and provides heat to the heap/stockpile leaching operations and the associated SX circuits. The flashed slurry will be cooled in an evaporative condenser and the solution will be decanted in a thickener. The residue will be filtered, repulped and washed in a three-stage CCD circuit. The washed residue will be neutralized with milk-of-lime and pumped to a suitably designed tailings impoundment.
Ancillary equipment such as the cooling water system, reagent preparation and grinding media system will also be provided. The oxygen will be supplied through an "over-the-fence" contract, with the oxygen plant owned and operated by a third party.
The pressure-leach vessels are housed in a partially enclosed building to protect the vessel shells and brickwork from excessive differential temperatures. The remainder of the equipment will be located outdoors. Electrical substation, motor control centers (MCCs) and a dedicated control room will also be provided. Superfine grinding equipment selection. B ased on the experience at Bagdad with the vertical stirred mills (i.e., two Metso 355 kW SMD mills in series) and the consistent inability to meet the P^sub 98^ specification in that circuit due to coarse particle carryover, a trade-off study was completed to evaluate various mill options for the Morenci application. This study resulted in the selection of a single IsaMill M10000 fitted with a 2.6 MW drive motor from Xstrata Technology (Brisbane, Australia) for Morenci. Although it was possible that the Metso SMD option could meet the design criteria at Morenci with two lines of three mills configured in series (a total of six mills), the experience at Bagdad seemed to indicate that even this configuration would be unlikely to meet the P^sub 98^ specification on a consistent basis (Williams et al., 2007). Finally, the IsaMill was tested at Bagdad at a pilot scale in a direct (head-to-head) comparison with the vertical stirred mills, and this yielded positive confirmation that the IsaMill could provide the required performance in this application.
Copper recovery and scale-up assumptions. Based on the performance of the Bagdad demonstration plant in both hightemperature and medium-temperature modes of operation during 2003 to 2005, there are a number of factors related to scale up that are known to affect copper extraction. These are summarized as follows:
* method of superfine grinding (near plug-flow vs. stirred tank equipment, and the grinding circuit configuration);
* the volume-flow characteristics of large-scale vs. small-scale pressure-leach vessels, which results in less bypass flow at commercial scale for a well-designed system;
* the presence of excess flotation reagents in concentrate samples (as a result of concentrate preparation at a small scale); and
* solution chemistry optimization (difficulties at small pilot scale vs. commercial scale).
Table 5 - Morenci process design criteria.
Morenci will utilize the IsaMill M10000 horizontally stirred mill to control the P^sub 98^ of the product closely. Based on the expected improvement in feed size distribution, coupled with consideration of the other factors listed above, overall copper recovery at Morenci is expected to be more than 97%. However, this will depend on the optimization of the flotation circuit (particle size and mineralogy of concentrate, amount and type of residual reagents present), the performance of the IsaMill M10000 and the resulting concentrate particle size distribution, and the optimization of solution chemistry in the pressure-leaching circuit. Recently completed batch-scale pressure-oxidation tests using actual Morenci concentrate at Hazen Research support the conclusion that >97% extraction is achievable. Further continuous pilot runs are planned to confirm the performance of the process using the expected feed material.
In addition to the improved particle size distribution (i.e., P^sub 98^ and P^sub 80^) expected with the Isa Mill M10000, Lightnin A-340 mixers will be utilized in all compartments of the pressure- leach vessel. Improved bulk fluid mixing performance attributed to the A-340 agitators has been demonstrated at Bagdad, and this is thought to greatly enhance the stability of the process by providing a more uniform reaction temperature than that achieved with Rushton Turbines. A review of the performance of the A-340 in the Bagdad pressure-leach vessel has been presented previously (Gigas and Wilmot, 2006).
Morenci project schedule. Detailed engineering commenced in June 2005 and construction began in March 2006. The first pressure-leach vessel was delivered to the site in November 2006, and the second vessel arrived in December 2006. Completion of the construction is scheduled for the third quarter of 2007.
Morenci capital and operating costs. The capital cost for the concentrate pressure-leaching facility was estimated to be $109 million, based on the June 2005 feasibility study. Also based on the June 2005 feasibility study, the cash operating cost, net of acid credits, was expected to be $0.14/lb cathode copper produced. On this basis, when compared with internal transportation, smelting and refining of the concentrate (at Miami, Arizona), this technology is expected to provide the Morenci operation with cash and full cost savings of approximately $0.15/lb and $0.08/lb, respectively. In this application, the technology provides an acceptable return on investment down to equivalent TCRCs of $75 per metric ton and $0.075/ lb. However, it should be noted that no new SX or EW capacity was required to be installed for the Morenci application.
Summary and conclusions
Continuous pilot-scale medium-temperature pressure-leach testing of surrogate concentrate samples, representing three main ore types from the Western Copper deposit in Morenci, Arizona, was conducted at Hazen Research. Aprocess flowsheet was developed for the commercial application of the MT-DEWSX process to the Morenci- Western Copper application. In parallel with this development work, a similar MT-DEW-SX flowsheet was demonstrated at a large scale in Bagdad, Arizona, and a fatal-flaw analysis was completed, preparing the way for a larger-scale commercial application.
Based on the above, detailed design for the application of the MT- DEW-SX process to treat flotation concentrates from the Western Copper district in Morenci, Arizona, was completed and construction was initiated in 2006. The process includes the use of an IsaMill M10000 unit for superfine grinding of the concentrate, two4.5-m- (14.8-ft-) diameterby 25.2-m- (82.7-ft-) long pressure leach vessels, single-stage flash pressure let down, decantation and filtration to generate strong PLS and countercurrent decantation (CCD) to generate weak PLS. The strong PLS will be fed directly to electrowinning for copper recovery. The weak PLS and lean electrolyte bleed will be blended with heap/stockpile leach PLS and processed through the existing SX circuit for copper recovery. Some byproduct sulfuric acid, which is utilized in the heap/stockpile leaching operations at Morenci, will be generated. No new or additional SX or EW facilities are required to be constructed for this project.
The Morenci MT-DEW-SX facility is scheduled to start up in the third quarter of 2007. The capital cost of the facility is approximately $109 million. At full operation it is expected to provide cash and full cost savings of $0.15/lb and $0.08/lb, respectively, compared with shipment of the concentrate to a smelter/ refinery complex.
The MT-DEW-SX process provides a safe, environmentally sound and economically attractive alternative to smelting and refining of concentrates in the Morenci Western Copper application. It is expected to provide similar benefits in other selected applications, for example, to treat concentrates that do not have smelting capacity available locally or regionally; to process materials containing significant levels of problematic contaminants, such as arsenic and antimony; and to use in situations where sulfuric acid is required for heap, stockpile and/or agitated leaching operations. A feasibility study was recently completed for the application of this technology to treat concentrate from Cerro Verde in Peru, and the options are under evaluation. Other potential applications of this technology are under consideration.
The authors thank Freeport-McMoRan Copper & Gold Inc. for permission to publish this paper. Many Freeport-McMoRan and heritage Phelps Dodge staff members contributed to the concentrate-leaching developments described in this paper, and in particular, the efforts of the hydrometallurgical staff at Bagdad and Morenci are recognized. The authors also thank Doug Halbe for his help with the analysis and design of the superfine grinding circuit, the staff of Hazen Research for their assistance with metallurgical test work and process development and Aker-Kvaerner Metals Division in Tucson for their assistance with design, procurement and construction of the facilities at Morenci.
Paper number MMP-07-030. Original manuscript submitted online July 2007 and accepted for publication August 2007. Discussion of this peer-reviewed and approved paper is invited and must be submitted to SME Publications Dept. prior to May 31, 2008. Copyright 2007, Society for Mining, Metallurgy and Exploration Inc.
Brewer, R.E., 2004, "Copper concentrate pressure leaching - plant scale-up from continuous lab testing," Minerals & Metallurgical Processing, November, Vol. 21, No. 4, pp. 202-208.
Gigas, B., and Wilmot, J.C., 2006 "Computational analysis and commercial demonstration of improved pressure leach vessel agitator design," ALTA 2006 Copper-10 Technical Proceedings, ALTA Metallurgical Services, Castlemaine, VIC, Australia.
Marsden, J.O., Brewer, R.E., and Hazen, N., 2003, "Copper concentrate leaching developments by Phelps Dodge Corporation," Hydrometallurgy 2003: Proceedings of the 5th International Symposium in Honor of Professor Ian Ritchie, Volume 2, Electrometallurgy and Environmental Hydrometallurgy, C.A. Young, A.M. Alfantazi, C.G. Anderson, D.B. Dreisinger, B. Harris and A. James, eds., The Metallurgical Society of AIME, pp. 1429-1446.
Marsden, J.O., and Brewer, R.E., 2003, " Hydrometallurgical Processing of Copper Concentrates by Phelps Dodge at the Bagdad Mine in Arizona," ALTA 2003 Copper-8 Technical Proceedings, ALTA Metallurgical Services, Castlemaine, VIC, Australia.
Marsden, J.O., Wilmot, J.C., and Hazen, N., 2007a, "Medium- temperature pressure leaching of copper concentrates - Part I: Chemistry and initial process development," Minerals & Metallurgical Processing, Vol. 24, No. 4, November, pp. 193-204, SME, Littleton, Colorado.
Marsden, J.O., Wilmot, J.C., and Hazen, N., 2007b," Medium- temperature pressure leaching of copper concentrates-Part II: Developmentof direct electrowinning and an acid autogenous process," Minerals & Metallurgical Processing, Vol. 24, No. 4, November, pp. 205-217, SME, Littleton, Colorado. Marsden, J.O., Wilmot, J.C., and Hazen, N., 2007c, "Medium-temperature pressure leaching of copper concentrates - Part III: Commercial demonstration at Bagdad, Arizona," Minerals & Metallurgical Processing, Vol. 24, No. 4, November, pp. 218-225, SME, Littleton, Colorado.
Williams, S" Gillaspie, J.D., Mathern, D.R., and Wilmot, J.C., 2007, "Super-fine grinding of concentrates for pressure leach processing by Phelps Dodge, Proceedings of Copper/Cobre 2007, Toronto, Canada. Riveras, P., etal., eds., CIM, Banff, Alberta, Canada.
Wilmot, J.C., Smith, R.J., and Brewer, R.E., 2004, "Start up and operation of the Phelps Dodge concentrate leach facility at the Bagdad Mine in Arizona," Hydro-Sulfides 2004 - International Colloquium on Hydrometallurgical Processing of Copper Concentrates, April 16-19, 2004, Santiago, Chile.
Wilmot, J.C., Smith, R.J., and Brewer, R.E., 2004, "Concentrate leach plant startup, operation and optimization at the Phelps Dodge Bagdad Mine in Arizona," Pressure Hydrometallurgy 2004 - International Conference on the Use of Pressure Vessels for Metals Extraction and Recovery, M.J. Collins and V.G. Papangelakis, eds., Banff, Alberta, Canada, October, 2004, Canadian Institute of Mining, Metallurgy and Petroleum.
J.O. Marsden and J.C. Wilmot
Senior vice president, technology and product development, and manager, concentrate leaching, respectively, Freeport-McMoRan Copper & Gold Inc., Phoenix, Arizona
Manager, Concentrate Leach Plant, Freeport-McMoRan Copper & Gold Inc., Morenci, Arizona
Copyright Society for Mining, Metallurgy, and Exploration, Inc. Nov 2007
(c) 2007 Minerals & Metallurgical Processing. Provided by ProQuest Information and Learning. All rights Reserved.